Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Wikidata

wikidata.org

Post reply on HN
Print view View original

61–69 of 69 posts

Re: Wikidata

#61

Post by arp242 »

bawolff wrote: post #60
arp242 wrote: post #52

Earlier quoted context omitted.

I found SPARQL pretty hard to deal with. Some time ago I wanted to make an overview of various election results to prove a certain point. The data is in there, I can see it in the UI, but I had a lot of trouble writing a script to retrieve it (so I could make a nice table and/or chart). Basically, after an hour I failed and gave up. I will readily admit this is due to my own ignorance; I just wanted to post a point i…

I think a big reason that sparql is hard, is it looks like SQL but has a really different data model. People who know sql tend to apply sql intuitions to sparql, which leads to confusion. But once you get past the learning curve its really not that bad.

> once you get past the learning curve its really not that bad.

Kind of applies to everything, no? :-)

But yeah, I don't think GraphQL is "bad", it's just hard to get started with (same applies to SQL really, although we're all used to that so we've forgotten about getting started with it). I want to use WikiData maybe ... once a year? Or less? I will probably have forgotten a lot of stuff next time I want to use it.

Re: Wikidata

#62

Post by kzrdude »

thom wrote: post #54
arp242 wrote: post #52

Earlier quoted context omitted.

I found SPARQL pretty hard to deal with. Some time ago I wanted to make an overview of various election results to prove a certain point. The data is in there, I can see it in the UI, but I had a lot of trouble writing a script to retrieve it (so I could make a nice table and/or chart). Basically, after an hour I failed and gave up. I will readily admit this is due to my own ignorance; I just wanted to post a point i…

Yeah, happy to admit SPARQL itself is hard, and every query requires having five wikidata tabs and three stackoverflow ones open. But, I find myself battling through because the results are worth it.

How do you deal with incomplete data?

I made a script to get the highest mountains in norway and get them on a map in Jupyter - looks nice, includes pictures if there are etc, but it doesn't pick out all the highest mountains. Most of them, but some are missing.

Re: Wikidata

#63

Post by kzrdude »

zaik wrote: post #32
HelloNurse wrote: post #28

Tests are a bit depressing. - "Borane" has two unlinked entries: Q127611 is "any chemical compound composed of boron and hydrogen atoms only" while Q15634214 is specifically boron trihydride. Both correct, but the latter should be labelled as an instance of the former. - "Anomalocaris" has Q37395 for the "extinct genus of radiodon" (instance of "fossil taxon") but species (e.g. Q49557506 Anomalocaris cranbrookensis)…

> for example Q936518 ("aerospace manufacturer") doesn't have a list of its instances You can query the database using SPARQL: https://w.wiki/3XgD

Some kind of property path was used, info here: https://en.wikibooks.org/wiki/SPARQL/Property_paths

Re: Wikidata

#64

Post by mmarx »

aasasd wrote: post #49
M2Ys4U wrote: post #45

Earlier quoted context omitted.

Wikipedia infoboxes can (and in small numbers are) being populated by Wikidata information. Some languages are making more use of this than others.

I'm talking about transfer in the other direction, or bidirectional one. Because people fill the infoboxes way more than they do Wikidata.

A bidirectional integration, Wikidata Bridge[0], is currently under development.

[0] https://www.mediawiki.org/wiki/Wikidata_Bridge

Re: Wikidata

#65

Post by thom »

kzrdude wrote: post #62
thom wrote: post #54

Earlier quoted context omitted.

Yeah, happy to admit SPARQL itself is hard, and every query requires having five wikidata tabs and three stackoverflow ones open. But, I find myself battling through because the results are worth it.

How do you deal with incomplete data? I made a script to get the highest mountains in norway and get them on a map in Jupyter - looks nice, includes pictures if there are etc, but it doesn't pick out all the highest mountains. Most of them, but some are missing.

I will admit I rarely use it for things that are so mission critical that this would matter. That said, it's always nice to leave the data more complete than you found it in these cases.

Re: Wikidata

#66

Post by zvr »

leoxv wrote: post #34
geraltofrivia wrote: post #26

Earlier quoted context omitted.

Conzept looks really good! Are there some whitepapers (apart from the user guide) that I can peruse? Some information about the team? Is this an LOD based solution where we can plug in arbitrary KGs? Did you fork the UI from some existing Knowledge Graph viz tools, or is it made from scratch?

Hey, good to know you like it! Its just me (Jama Poulsen) currently. I hope at some point to be able to work with more people on this project and also to have a base for other archives / institutions (as a B2B product and service) to use the "Conzept UI" for their knowledge base. The Conzept framework is already pretty generalized, multi-lingual and customizable, but more needs to be done. I'm aiming to be able to do…

Congrats on building this!

Is the source under an Open-Source license? Can it be seen anywhere? Do you accept contributions?

Re: Wikidata

#67

Post by aasasd »

mmarx wrote: post #64
aasasd wrote: post #49

Earlier quoted context omitted.

I'm talking about transfer in the other direction, or bidirectional one. Because people fill the infoboxes way more than they do Wikidata.

A bidirectional integration, Wikidata Bridge[0], is currently under development. [0] https://www.mediawiki.org/wiki/Wikidata_Bridge

Excellent!

Re: Wikidata

#68

Post by leoxv »

zvr wrote: post #66
leoxv wrote: post #34

Earlier quoted context omitted.

Hey, good to know you like it! Its just me (Jama Poulsen) currently. I hope at some point to be able to work with more people on this project and also to have a base for other archives / institutions (as a B2B product and service) to use the "Conzept UI" for their knowledge base. The Conzept framework is already pretty generalized, multi-lingual and customizable, but more needs to be done. I'm aiming to be able to do…

Congrats on building this! Is the source under an Open-Source license? Can it be seen anywhere? Do you accept contributions?

It's not opensource. I'm open for feature requests or creating new API/site integrations. If you are missing something, please let me know.

Re: Wikidata

#69

Post by chris_wot »

geraltofrivia wrote: post #23
chris_wot wrote: post #14

I have frequently created articles and had Wikidata linked to the article. Take https://en.wikipedia.org/wiki/Kate_Baker - if you click on Wikidata item, then it takes you to: https://m.wikidata.org/wiki/Q21536118

Thank you for doing that!

A pleasure, researching articles is always fascinating.
Post reply on HN